US 12,276,589 B1
Method and system for detecting oil abrasive particle size
Xinyu Pang, Taiyuan (CN); Shijie Cui, Taiyuan (CN); Mingyuan Guo, Taiyuan (CN); Kaibo Lv, Taiyuan (CN); Feng Li, Taiyuan (CN); Xuewen Wang, Taiyuan (CN); Yan Wu, Taiyuan (CN); Yong'ai Cao, Taiyuan (CN); Pengfeng Shao, Taiyuan (CN); Yuewei Wang, Taiyuan (CN); Jiawei Bai, Taiyuan (CN); and Yixiang He, Taiyuan (CN)
Assigned to Taiyuan University of Technology, Taiyuan (CN); and Shanxi KEDA AUTOMATION CONTROL co., Ltd., Taiyuan (CN)
Filed by Taiyuan University of Technology, Taiyuan (CN); and Shanxi KEDA AUTOMATION CONTROL Co., Ltd., Taiyuan (CN)
Filed on Nov. 27, 2024, as Appl. No. 18/961,577.
Claims priority of application No. 202311598217.2 (CN), filed on Nov. 27, 2023.
Int. Cl. G01N 15/02 (2024.01); G01N 33/28 (2006.01)
CPC G01N 15/0266 (2013.01) [G01N 33/2858 (2013.01)] 8 Claims
OG exemplary drawing
 
1. A method for detecting an oil abrasive particle size, comprising:
acquiring a voltage signal of an oil abrasive particle in real time, wherein the voltage signal is generated in real time in a process that the oil abrasive particle passes through an electromagnetic abrasive particle sensor, specifically comprising:
for a multimode arrayed electromagnetic abrasive particle sensor arranged in a square shape or circumferentially, obtaining a center of symmetry B of a combined shape of all channels according to sections perpendicular to central axes of the channels; classifying all channels of which the central axes have an equal distance to the center of symmetry B as channels of a same class; classifying multimode arrayed channels corresponding to the channels of the same class as a group of channels, wherein channels in a same group of channels have a consistent inner diameter size; selecting any channel in any group of channels, placing oil abrasive particles that have different sizes at the central axis of the channel for uniform linear motion, and putting abrasive particles of a same size in uniform linear motion once, thereby generating different voltage signals;
establishing a voltage curve coordinate system with an amplitude of the voltage signal as a Y-axis and a time of the voltage signal as an X-axis, and obtaining a voltage curve;
performing feature extraction on the voltage curve, and obtaining an area S defined by a voltage curve within a half voltage signal cycle T and the X-axis and a maximum absolute value A of a voltage signal value according to the voltage curve coordinate system; and
inputting the area S and the maximum absolute value A of the voltage signal value to a trained back propagation (BP) neural network for prediction to obtain a predicted oil abrasive particle size.